AN INVESTIGATION OF TREE-RING RESPONSE TO EXTREME FLOOD EVENTS ALONG THE SCHOHARIE CREEK, CENTRAL NEW YORK
Tree ring widths in each core were measured and analyzed using the ANTEVS time-series analysis program (Vollmer, 2014). Ring sequences were detrended using a 10 year spline to remove longer wave-length growth signals unrelated to flooding events, i.e.: natural lifetime variability and changes in canopy cover. Four trees which did not have multiple cores achieving a cross-correlation of R > 0.60 were removed from the data set. Averaged sequences for each of the two sites were then generated and found to cross-correlate at R=0.43. Although the trees at one site are about 2 m higher above creek bankfull level, there appears to be significant correlation between the two sites. A master sequence constructed from all accepted trees provides annual ring growth trends from 1907-2013. We then compared ring growth to annual (water year) peak discharge from 1927-2013 using ANTEVS. There was no statistically significant correlation.
Visually, larger floods (>1400 cms) appear to coincide with diminished ring growth, however decreased tree ring widths can be seen in both low discharge years as well as extreme flood years. Less extreme flood years appear to correlate positively with ring width. At this point, it is not possible to differentiate between small rings caused by drought vs. floods, but a planned comparison between floodplain and upland trees may help to tease this apart.